IZA DP No. 3029

An Experimental Investigation ofAge Discrimination in the English Labor Market

Peter A. RiachJudith Rich

DI

SC

US

SI

ON

PA

PE

R S

ER

IE

S

Forschungsinstitutzur Zukunft der ArbeitInstitute for the Studyof Labor

September 2007

An Experimental Investigation of Age

Discrimination in the English Labor Market

Peter A. Riach IZA (Research Fellow)

Judith Rich

University of Portsmouth and IZA

Discussion Paper No. 3029 September 2007

IZA

P.O. Box 7240 53072 Bonn

Germany

Phone: +49-228-3894-0 Fax: +49-228-3894-180

E-mail: iza@iza.org

Any opinions expressed here are those of the author(s) and not those of the institute. Research disseminated by IZA may include views on policy, but the institute itself takes no institutional policy positions. The Institute for the Study of Labor (IZA) in Bonn is a local and virtual international research center and a place of communication between science, politics and business. IZA is an independent nonprofit company supported by Deutsche Post World Net. The center is associated with the University of Bonn and offers a stimulating research environment through its research networks, research support, and visitors and doctoral programs. IZA engages in (i) original and internationally competitive research in all fields of labor economics, (ii) development of policy concepts, and (iii) dissemination of research results and concepts to the interested public. IZA Discussion Papers often represent preliminary work and are circulated to encourage discussion. Citation of such a paper should account for its provisional character. A revised version may be available directly from the author.

IZA Discussion Paper No. 3029 September 2007

ABSTRACT

An Experimental Investigation of Age Discrimination in the English Labor Market

Carefully-matched pairs of written job applications were made to test for age discrimination in hiring. A twenty-one year-old and a thirty-nine year-old woman applied for jobs where a “new graduate” was sought; men aged twenty-seven and forty-seven, inquired about employment as waiters; women aged twenty-seven and forty-seven, inquired about employment in retail sales. The rate of net discrimination against the older graduate, and against the older waiters in their London inquiries, correspond to the highest rates ever recorded anywhere, by written tests, for racial discrimination. There was a statistically significant preference for the older applicant in retail sales. JEL Classification: J71, C93 Keywords: age, discrimination, employment, field experiment, hiring Corresponding author: Judith Rich Department of Economics University of Portsmouth Richmond Building Portland Street Portsmouth PO1 3DE United Kingdom E-mail: judy.rich@port.ac.uk

“He looks old to be a waiter” (White Teeth)

I. Introduction

The growing interest in field experiments, as a method of empirical investigation in

economics, is demonstrated by the special issue of The BE Journal of Economic

Analysis and Policy (Advances), (2006, 6, issue 2) which is dedicated to field

experiments. The first field experiment of discrimination in employment, which used

pairs of matched, written job applications, was undertaken in the 1960s by Jowell and

Prescott-Clarke (1970). They developed the technique to investigate racial

discrimination in employment in England. The first time this experimental method

was applied to investigate sexual discrimination in employment was in Melbourne

during the 1980s (Riach and Rich 1987). The measurement of discrimination by

making matched, written job applications has been received with approbation in the

academic journals, for example - in a survey of evidence on discrimination, Darity and

Mason stated “This (correspondence testing) is impressive direct evidence of

discrimination from a powerful test procedure” (Darity and Mason 1998, p. 81).

Although it originated almost forty years ago, there has been virtually no application

of this technique to measuring the extent of age discrimination in employment. This is

surprising, given the widespread contemporary concern about the economic

implications of the ageing population in Western countries, and the frequent

recommendation that the only viable solution to the consequent “pensions’ crisis” is

an extended working life. For example the Organisation for Economic Co-operation

and Development (OECD), referring to the United Kingdom - “… in 2050, for every

person over the age of 65, there will be only 2.1 individuals of working age compared

to 3.7 in 2003” (OECD 2004, p. 36). The British Government’s response to this

impending demographic difficulty came in a Green Paper issued in 2002 by the

Department of Work and Pensions; Simplicity, security and choice: Working and

saving for retirement. Amongst other observations it included; “Increasing

employment among older workers is essential if we are to address the pensions

challenge. Working longer can dramatically reduce the rate at which people need to

save for their retirement” (Department for Work and Pensions 2002, Cmd. 5677, p. 93

- emphasis added).

It is the case, though, that there is a significant challenge in applying the experimental

technique to age discrimination; this is the obvious variation in human capital across

the generations. The logic of this experimental technique, as innovated by Jowell and

Prescott Clarke, is to design the test so as to control strictly for human capital

components such as education, qualifications, skills and experience, and so that the

only distinguishing feature of the two job applicants is the characteristic, such as race

or sex, which is being tested. The influence of race or sex on hiring decisions is

consequently isolated. In the case of age there must inevitably be a variation in the job

experience of the different age groups, and therefore a difficulty in determining

whether any employment preference is attributable to a profit-maximizing response to

differential human capital or to prejudice. The one economist who has applied the

technique to age, Bendick (1996 and 1999), attempted to deal with this issue by

having older applicants who had spent 25 years in some unrelated activity, such as

child-raising, military service or public school teaching, which generated no relevant

experience for the employment being tested. This is a highly artificial construct and

leads to uncertainties; for instance, do employers rate experience between ages 40 and

2

50 as equivalent to experience between ages 25 and 35? Also it is not a realistic basis

for policy development as virtually no older workers fit such a pattern.

In our paper “Field Experiments of Discrimination in the Market-Place” (Riach and

Rich 2002) we recommended that instead of adapting job applicants to the technique,

the technique should be adapted to the special case of older applicants. In other words,

to accept that the job experience component of human capital does vary between

different groups and have realistic candidates make applications, but to control for all

other dimensions of human capital.

A frequent accusation against older applicants is that they are less mentally

able/flexible and less physically active than their youthful competitors; “… numerous

surveys and research conducted in the past 15 years point to negative employer

perceptions vis-à-vis older workers with respect to their productivity, cost, work

motivation, health, receptiveness towards training and ability to cope with

technological and organisational change” (OECD 2004, p. 97; see also Purcell et al.

2003, pp. 3-4). However the objective scientific literature is to the contrary; “The

finding from more than 100 research investigations is that there is no significant

difference between the job performance of older and younger workers” (Warr 1994, p.

309).

We decided to confront such ageist attitudes by presenting older applicants who were

not more than forty-seven, who were engaged in strenuous physical activity, such as

competitive squash and cycling, and who demonstrated mental flexibility by an up-to-

date interest in computers and information technology. In other words we controlled

3

for the older candidate’s mental and physical capacities, but not for their length of

experience. In which case, if a preference were found for younger applicants with

twenty years less experience, it would indicate a very significant level of prejudice

against older applicants. On the other hand, if we were to find a preference for the

older workers in such circumstances it could be interpreted as an economically

rational response to human capital superiority, rather than prejudice against youth.

The OECD recognises that; “…age discrimination is neither overt nor easily

measured” (OECD 2004, p. 98). This is a challenge which we now address; what

follows is the first realistic attempt to measure age discrimination by using the

experimental technique of forwarding matched, written applications.

2. The experiment

The intention was to have pairs of job applicants who were carefully-matched in all

respects except in the experience which inevitably goes with age. An implication of

this approach is that jobs with a career hierarchy were ruled out of the investigation.

For instance, academics in their mid-twenties would be applying for different posts to

those in their mid-forties: the former would be applying for post-doctoral fellowships

or lectureships, whilst the latter would be applying for Chairs or Deanships. This is

not to say that age discrimination may not be alive and well in academia, or in law, or

in the civil service, but instead that it cannot be investigated by the technique of paired

mail applications. We have chosen occupations where it is realistic to expect that

applications will come from candidates aged twenty years apart.

There are two techniques for applying this experimental method. The first is to

respond to advertized vacancies, as innovated by Jowell and Prescott-Clarke in 1969,

4

and surveyed by Riach and Rich (2002). The second is to conduct the investigation by

forwarding unsolicited job inquiries to a group of employers in some occupation, as

innovated by Fidell in 1970 and surveyed by Riach and Rich (2004b). The former

procedure confronts the employer with simultaneous pairs of fictitious job

applications at a time when he/she has initiated recruitment and they will be dealt with

during the normal hiring process, but the application and observation rate, and

therefore the duration of the study, is dependent on the state of the labor market. The

latter procedure is more expedient; all that is required is an appropriate listing of

employers in a particular occupation, and two mailings about a month apart, if the

intention is to test matched pairs and if suspicion on the part of employers is to be

allayed. This procedure is more appropriate in occupations where inquiries are

customarily initiated from the supply side and in small business, so ensuring the

inquiry is likely to be answered by the same person who normally takes hiring

decisions. As inquiries must arrive at least a month apart there is some randomness in

the extent to which they will coincide with a vacancy, therefore there can be less

expectation of “equivalent treatment” (both candidates being offered interviews) than

with the first procedure. Nevertheless, if in the first mailing fifty per cent of inquiries

go from candidate A and fifty per cent from candidate B, with reversal in the

following month, this variation in timing will not bias the overall outcome of the

experiment: i.e. there is control by age for the timing of application receipt.

We decided to apply both techniques in this study. In England we applied to

advertized vacancies for new graduates in those positions where a “degree in general”

was the prerequisite for employment, rather than any specific degree. Vacancies were

obtained from the Saturday Guardian and from various web-sites

5

(www.jobs.guardian.co.uk; www.topjobs.co.uk; www.monster.co.uk). Entry into the

profession of Chartered Accountancy is by application to an authorised training firm,

consequently such firms expect to receive, without advertisement, a steady flow of

inquiries. We also had our new graduates apply to all such firms in England. One of

our new graduates was female and aged twenty-one; the other was female and aged

thirty-nine i.e. an individual euphemistically designated in Britain as a “mature age”

graduate. The latter applicant had worked for eleven years as a secretary and spent

five years in full-time child-rearing before entering university. She was divorced with

a child of ten to indicate that pregnancy was likely to be a thing of the past.

We chose this area of employment because of its contemporary relevance; in 2002/03

there were 95,590 female students over thirty in English universities (20.4 percent of

total students) and the universities actively recruit this age group (Higher Education

Statistics Agency 2002/03, unpublished data). An additional reason for targeting a 39

year-old applicant was to explore the possibility of obtaining some experimental

confirmation of the survey evidence which suggests that age discrimination can

impact at a quite early age; e.g. “In a survey of over 1000 people the Chartered

Institute of Personnel and Development found that … 1 in 4 think that employers are

not interested in employing people over age 40” (Third Age Employment Network

2003). In 1996 a survey of employers found evidence that age discrimination started at

42 (Penna Sanders and Sidney 2002).

In an experimental investigation of age discrimination, for the reasons discussed

above, it is not possible strictly to alternate the résumés, as is done in race and sex

experiments, but in all non age-related characteristics the résumés of “new graduates”

6

were alternated. In half the applications the older applicant was a law graduate from X

and in the other half she was an economics graduate from Y. We could not use the

names of genuine educational institutions or employers in the résumés for two

reasons. First there is the risk of detection if an employer were to make direct contact

with a genuine company or university falsely cited in a résumé. Secondly an

educational institution might take legal redress against a party falsely claiming to

possess one of its awards. We decided to deal with this difficulty by inventing

fictitious universities and employers. Just such an approach had been adopted in the

International Labour Office’s investigation of racial discrimination in the German

labor market: fictitious schools and universities were invented (Goldberg et al. 1996).

There are approximately one hundred universities in Britain and all but a handful have

locational names; either of a city or a county. We therefore chose an English city and

county which did not have universities, but which quite plausibly might, and used

them in the résumés of our “new graduates”. There was a flood of new universities in

the decade prior to these tests. In 1992 twenty polytechnics became universities with

names like De Montfort, South Bank, Liverpool John Moores and London Guildhall.

Since then there has been a steady trickle of additions with Chichester, Southampton

Solent and Thames Valley amongst those acquiring universities. If counties such as

Hertfordshire and Staffordshire have universities why might not Herefordshire and

Shropshire? If towns like Loughborough, Bournemouth and Brighton have

universities is it not conceivable that Ipswich and Salisbury have universities? We

therefore believe that employers recruiting graduates would be unlikely to have a

definitive knowledge of the current list of universities. In the case of current and

former employers we simply specified their field of activity, such as merchant

banking or chartered surveying. University careers officers and a Course Leader in

7

Business Studies provided advice on the personal profiles used for the graduates’

résumés, and their realism and efficacy is confirmed by the fact that we did obtain

responses from very large firms and major recruitment agencies.

Cultural and sporting interests were chosen to be interchangeable because of the

practice of reversing résumés. They were also deliberately chosen to suggest that the

older applicant had no stereo-typical ageist traits; the interchangeable interests

included classical and contemporary dance, playing squash and language classes.

We sent unsolicited inquiries about possible job openings for male waiters to four

hundred and seventy hotels and restaurants throughout England (two hundred and

twenty in London and two hundred and fifty across the rest of the country). One

applicant was twenty-seven and the other was forty-seven. We chose this area of

employment because it is one where supply-side inquiries are customary, and because

it is an area of small business where any inquiry is likely to find its way to those who

normally take the hiring decision.

It was not possible to perform any reversal of résumés in this case as, unlike our “new

graduate” applicants, the education of waiters could not have been undertaken

simultaneously. Both candidates had completed year eleven of school, but in the case

of the older candidate the prevailing award in England was General Certificate of

Education (GCE) at Ordinary (O) level, whereas in the case of the younger candidate

it was General Certificate of Secondary Education (GCSE). Both candidates had

included English and Mathematics in their awards. In this case we invented fictitious

restaurant names for current employers. The efficacy of this tactic in particular, and

8

the content of our résumés in general is confirmed by the receipt of positive responses

from some of London’s most expensive and fashionable restaurants. “Interests” were

chosen to demonstrate the older applicant’s physical fitness, and capacity to cope with

modern technology. The “interests” included competitive squash and internet usage;

also, computing had been studied at evening class. The résumés of waiters were

prepared in conjunction with advice from the head waiter of a two-star restaurant in

London. In any one posting half the inquiries went from the older applicant and half

from the younger, with reversal in the following month. The résumés used for the

waiters are included in the Appendix.

We used an identical procedure in making inquiries about possible employment in

three hundred female clothing stores in London In this case one female applicant was

twenty-seven and the other forty-seven; fictitious names were invented for the retail

shops where they currently worked as assistant mangers. We chose to include this area

of employment because it is one of small business, where the inquiry is likely to be

dealt with by the person who normally takes hiring decisions. An additional reason is

that there is anecdotal evidence in England that retail sales is an area where some

employers do deliberately target older workers. The large “do-it-yourself” retailer,

B&Q, is noted for hiring older employees because of the benefit which their

experience provides customers.

Once again, the relevant content of our résumé in general is confirmed by the receipt

of positive responses from some of London’s more expensive and fashionable retail

shops and from major High Street chains. A former senior personnel manager of a

major retail chain advised us on the résumés for this occupation. In this case the

9

“Interests” chosen to demonstrate the older candidate’s physical fitness and mental

agility were competitive tennis, the internet and learning Italian. Computer usage had

been pursued at evening class.

The three pairs of English résumés were also vetted by an experienced employment

consultant who specialises in advising older job applicants. We applied by surface

mail and cited both an email and postal address for responses. The postal addresses

were in comparable socio-economic districts of central London, approximately one

mile apart. We have always been very careful to retain documentation of our research,

so we were careful to print off applications and replies, so hard copies could be filed.

Positive responses via email or surface mail were dealt with promptly and courteously

with a reply explaining that alternative employment had already been secured.

The one publication where the ethical considerations involved in this deceptive

procedure are dealt with, and compared with research activity in psychology,

sociology and laboratory-experimental economics, is Riach and Rich (2004a). The

alternative, non-deceptive, techniques for measuring discrimination have encountered

difficulties. Surveys of attitudes towards target groups in the labor market are not

likely to produce honest and accurate responses, as demonstrated by La Piere’s classic

study. In 1934 he travelled through the USA with a Chinese couple and gained

admittance to all except one of 241 hotels and restaurants approached. In response to

questionnaires sent six months later to the same establishments, over 90 per cent

replied they would not accept Chinese guests. (La Piere 1934) The econometrician’s

application of regression analysis to published data to deduce discrimination,

pioneered by Blinder (1973) and Oaxaca (1973) has been subject to considerable

10

criticism, which revolves around the specification of the model and the choice of

independent variables; see for example (Gunderson 1989) On the other hand,

carefully-designed deceptive field experiments can provide an unequivocal measure of

discrimination.

3. The results

The outcome of this experiment is set out in Table 1 in a format which follows

McIntosh and Smith (1974, p. 13) and which has since been adopted in field

experiments across Europe; e.g. Brown and Gay (1985); Bovenkerk (1992, pp. 26, 31)

(see Riach and Rich 2002, pp. F486-F491). Column 4 shows the number of occasions

when one or both applicants received a favorable response; by post, telephone, fax or

email. This total is divided as follows: column 5 shows occasions when both received

favorable responses (equal treatment); column 6 shows occasions when only the

younger received a favorable response (discrimination against the older); and column

7 shows occasions when only the older received a favorable response (discrimination

against the younger). Column 8 is net discrimination; that is 7 minus 6, so that it is

positive when the older applicant encountered more discrimination than the younger

applicant. The statistical significance of any finding of net discrimination was

determined by the application of the chi-square test. The data were categorised as

accepted /rejected for two applicants in a 2*2 contingency table (Riach and Rich 2002,

pp. F493 – F496). A comparison with British experimental results for race and sex

discrimination is provided in Table 2. The rate of net discrimination against the older

graduate applicant of 59.6 percent is one of the highest ever recorded. The highest net

rate of discrimination ever previously recorded by the written experimental method

was 66.7 percent against Antilleans in France in 1977 (Riach and Rich 2002, Table 4,

11

p. F500). In view of the encouragement given by government for people to retrain,

and by universities for mature-age women to enrol, this result is quite disturbing. Our

older female graduate had an equivalent degree, no greater likelihood of pregnancy

and one might have thought that her life experience would have enhanced her human

capital and employability. On the contrary, we had one reply which explained “I

wanted to be honest with you, our client is looking for recent graduates who are

looking for their first job. You obviously have substantially more experience than this

and being honest I feel that it would be a waste of your time to take your application

any further. Sorry if this sounds harsh but we do believe in being honest with people”.

(The younger applicant received a positive response: information on the job, and a

request to complete a maths test and a questionnaire).

We report, in Table 3, the results for the sub-set of graduates, whose applications were

to employers who possessed the imprimatur of “Investors in People”. The net rate of

discrimination in the case of these employers was 46.2 percent, which was statistically

significant at the 0.01 level. The Home Page of this organization states; “Investors in

People Standard is a straightforward, proven framework for delivering business

improvement through people…” (Investors in People 2005). It also states on the page;

Recruitment and Selection Model-Any Questions? “The Recruitment and Selection

Model focuses specifically on good recruitment and selection practices, and how they

impact on performance”. Apparently some members of Investors in People believe

that “good recruitment and selection practice” involves discarding applications from

graduates because they have reached the age of 39.

12

The rate of net discrimination against the 47-year-old waiter in London of 68.2 per

cent is the highest ever recorded anywhere by the written experimental method. Table

2 indicates that McIntosh and Smith (1974) recorded a net rate of 30.0 percent for

West Indians and Jowell and Prescott-Clarke (1970) recorded a rate of 11.0 percent

for West Indians and 50.0 percent for Indians. Outside London discrimination was at

the much lower rate of 13.8 percent so that the net rate for England was 28.8 percent.

In the case of retail sales we found a preference for the older applicant, which was

statistically significant at the 0.05 level.

Although there are no other published experimental studies of age discrimination in

employment available for comparative purposes, List (2004) has detected age

discrimination in a product market experiment. He found that white males aged 60 and

over encountered discrimination when they participated as either buyers or sellers in

the sportscard market. This reinforces our finding that age discrimination is a serious

phenomenon and needs to be addressed as vigorously as racial and sexual

discrimination.

4. Interpretation

Warr (1994) presents a classification of job activity in four categories, based on the

relationship of performance to age. First are “age-impaired activities” in which there

is a negative relationship between age and performance; “… basic capacities are

exceeded to a greater extent for older people and experience cannot help. Tasks of that

kind include continuous rapid information processing and some forms of strenuous

physical activity.” (Warr 1994, p. 314) “... complex tasks, requiring a large number or

13

processing steps, are especially likely to be susceptible to cognitive slowing” (Warr

1994, p. 315). Second are “age-counteracted activities”, in which there is no

relationship between age and performance, because older people have strategies to

compensate for any decline in information processing skills or in physical capacity.

For example “…middle managers may learned to conserve their energy and time by

operating through day-to- day tactics which reduce cognitive and affective load”

(Warr 1994, p. 317). Third are “age-neutral activities”, in which there is no

relationship between age and performance, because the work is relatively

undemanding and routine “...primary memory is apparently unaffected by age; older

people are as able as their younger counterparts to hold in memory small amounts of

information that are being used in uncomplicated cognitive activities” (Warr 1994, p.

317). Fourth are “age-enhanced activities” in which performance improves with age,

because of the favorable impact of experience. Knowledge-based activity without time

pressure comes within this category. For example; “… in a study of an American

company’s sales staff older employees were rated much more positively than younger

ones in almost every respect” (Warr 1994, p. 316).

Warr’s framework provides an explanation for age-related employment preferences

which arise from age-related human capital differences. Becker and Arrow/Phelps

have provided the economic bases for employment preferences which arise from

discriminatory attitudes in the market place. Becker’s (1971) theory proposes that

customers, employers and/or current employees will sacrifice economic benefit, in

order to indulge a “taste” for eschewing contact with some perceived pariah group.

The Arrow/Phelps (1973/1972) hypothesis of “statistical discrimination” refers to the

incomplete information, which employers have of the productivity and work

14

characteristics of individual job applicants; this induces employers to resort to

generalisations about the employment characteristics of groups, as a screening device

to minimise the cost of information acquisition in the hiring process.

In the case of “new graduates”, where we found a very high rate of discrimination

against the older job applicant, it is not realistic to explain it by attribution to Warr’s

first category; the older graduate is 39, and has just completed a degree in a rigorous

discipline: economics or law. Also she engages in squash or contemporary dance, and

learns Italian. It is difficult to credit that discrimination is “statistical” in the sense

defined by Arrow (1973) and Phelps (1972). Both the “new graduate” women are in

the fertile age range and, theoretically, equally prone to employment interruption,

although the age of the older candidate’s child might indicate she had done with

planned child-bearing. The nature of this employment is that employee-customer

contact is minimal to moderate, so it is unlikely that employers are being driven by

any customer pressure; instead a clue might be found in the reply quoted above;

delayed entry into higher education and experience of work being viewed as a

disadvantage, as it might have imparted confidence, self-assurance and a degree of

independence which could make life difficult for low-level managers, i.e. a variant of

employer discrimination à la Becker (1971). Some confirmation of this hypothesis

was found when graduates were interviewed about their experience of job search; “…

some employers appear to have regarded mature graduates with suspicion, not only

unsure about where they might fit into an organisation, but also wary of their motives

for doing a degree and imputing character flaws in those who had not followed the

“normal” (in fact middle-class) educational route from secondary school into higher

education” (Purcell et al. 2003, p. 26).

15

Waiters clearly come into Warr’s category three, as the work is relatively

undemanding and routine; also recall that our older waiter plays competitive squash,

so his physical capacity to do the job can hardly be in doubt. The contrasting finding

between London and the rest of England rules out “statistical discrimination”; it is not

realistic to hypothesise that any constituents of incomplete information vary

geographically, in such a way as to activate this reaction. As long ago as 1933 George

Orwell in Down and Out in Paris and London, (1940, pp. 68-69) recognized the

strange symbiosis between waiter and diner. It certainly is an occupation with a

critical interaction between employee and customer, and in an economic activity

where repeat business is vital for commercial success. There is critical interaction, for

instance, between nurse and patient, but few of us are repeat customers of hospitals.

This might suggest particular insight in the casual aside quoted at the outset of this

paper, which comes from the television adaptation of Zadie Smith’s novel White

Teeth, that is, customer prejudice à la Becker.

In their econometric study, which used data from the Workplace Employment

Relations Survey, Daniel and Heywood found “…strong evidence for the role of

deferred compensation and internal labor markets as a negative predictor of hiring

older workers. This fits the hypothesis that efficient life-time incentive structures

require hiring younger workers and employing them when old, but not hiring older

workers.” (Daniel and Heywood 2007, p. 49). However these factors cannot explain

our results for graduates and waiters. Waiters do not operate in internal labor markets,

nor do they benefit from deferred compensation. Our older graduate is 39, whereas

Daniel and Heywood distinguish between those under 50, and those who are 50 or

over. With more than twenty years until retirement our mature age graduate hardly fits

16

the hypothesis that - “… the firm does not hire older workers because their shorter

employment horizon means they are less well motivated by delayed compensation”

(Daniel and Heywood 2007, p. 37).

The geographical variation between London and the rest of England is unprecedented

in this type of experimental research. It may be partly explained by differential

unemployment rates: for the period of this experiment (July-September 2004) the

Labour Force Survey records a rate of 7.2 percent for men in London, in contrast to a

rate of 4.4 percent for the rest of England (Labour Market Statistics 2005). When

unemployment is high, and more applicants are searching, it facilitates the exercise of

any discriminatory penchant which employers may have. Conversely when

unemployment is low, and fewer applicants are searching, employers have less

opportunity to discard applicants simply on the basis of some arbitrary characteristic

such as race, sex or age. The higher rate for London may also be partly attributable to

a greater devotion to pursuit of the fashionable “celebrity lifestyle” in the

cosmopolitan capital, with its emphasis on the “youth culture”, which reflects the

younger age distribution of the London population. The proportion of the Inner

London population aged 20-39 in 2001 was 41.3 percent; the proportion of the

population in that age range for the rest of England was 27.5 percent (Census 2001).

Retail managers clearly come into Warr’s category four and, significantly, it is sales

staff whom he cites as his example of age–enhanced activity. Both our applicants had

managerial backgrounds, and this is the most commercially-responsible job tested, so

we interpret the preference for the older applicant as reflecting a realisation by

employers that this is an employment where the job experience component of human

17

capital contributes significantly to performance, which may include customer

satisfaction. We have here a rational response to age-related human capital

differences, rather than discrimination.

5. Policy: implications and recommendations

We found net discrimination against a 39-year-old graduate of 59.6 per cent and

against a 47-year-old waiter, in London, of 68.2 percent. In the 1960s Political and

Economic Planning (PEP), recorded discrimination of 90.0 per cent against Indians

and Afro-Carribeans (Daniel 1968). The finding of that level of racial discrimination,

by PEP, gave rise to considerable concern at the time: not least in the House of

Commons during the Committee Stage and the Second Reading of the Race Relations

Bill in 1967/8. All three major parties (including the Conservatives, who are

ideologically committed to laissez faire) referred, with obvious concern to PEP’s

findings: Quentin Hogg the Conservative member for St. Marylebone (later Lord

Hailsham, the Lord Chancellor) said, during the Committee stage; “… a great deal has

happened in the last year. There has been the first Report of the Race Relations Board.

There has been the PEP Report on Racial Discrimination … I tell the right hon.

Gentleman plainly that, originally I was very critical of fresh legislation on this

subject so soon after the last. But I was immensely impressed by the quality of some

of the documents to which I have referred” (Parliamentary Debates 15 November

1967). Subsequently during the Second Reading Debate he said; “I was convinced in

the end by the evidence of the PEP report and the Street Report…that in the fields of

housing and employment there are circumstances in which the economic laws would

operate in favor of discrimination and against human rights” (Parliamentary Debates

23 April 1968).

18

It was right and proper that the British nation’s statesmen reacted so, and took

legislative action to confront racial discrimination in 1968. It is not unreasonable to

expect equivalent concern and reaction in 2007 in respect, not just of the elderly, but

also of the “young seniors” (for example our waiters), and, in the case of the graduate,

the middle-aged. They are more numerous and just as deserving of protection as the

black community. The portents, however, are not good; the deadline for implementing

a European Union Directive on age equality legislation, agreed in 2000, was

December 2006; the British legislation only came into force in October 2006.

An abiding mantra of British governments for the past twenty-five years has been the

need for a “flexible labor market”. It is rarely defined, but it is demand-side flexibility

which is always implied. One official definition is; “In a “flexible” labor market

where employment is little regulated (in terms of pay, working hours, restrictions on

dismissal etc.) the creation of low-paid, part-time, short-term or otherwise non-

standard jobs is unconstrained, and there is a high level of job turnover, employers

screen less intensively before hiring” (OECD 1992, p. 207).

In the Green Paper, Simplicity, Security and Choice: Working and Saving for

Retirement, the current British government has shown great concern for the

demographic structure of the labor force; for instance, it estimates that the ratio of

those 65 plus to those 15-64, will rise from 24.4 percent in 2000 to 32.8 percent in

2025 and to 39.2 percent in 2050 (Department for Work and Pensions 2002, p. 16). As

we saw in the Introduction, the Government considers it essential that employment

rates of older workers rise to meet this demographic/pensions crisis. It follows that a

government, which proselytises for demand-side labor market flexibility, and which

19

wants to encourage people to work longer, has an obligation to ensure supply-side

flexibility for older workers, so that they are not trapped in the unemployment pool, or

in unsatisfactory or oppressive current employment.

Experimental research (e.g. McIntosh and Smith; Riach and Rich 1987, 2006) of the

hiring process has repeatedly demonstrated the uninformative, and sometimes

dishonest, nature of rejection letters. The pattern of dishonesty was repeated in this

study. For example, in an application for a graduate position, the younger applicant

was sent the following response on 11 November:

“Thank you for sending your CV to …. Can you call me to discuss …”.

The older applicant was sent the following response on 19 November:

“I am writing to inform you that you have not been selected for a registration

interview with … on this occasion. Unfortunately your skills and experience to date

do not exactly match our client’s requirements at this time”.

On another occasion the older graduate was sent the following on 7 August:

“… I regret to advise you we have no vacancy for a trainee chartered accountant. We

have now filled all our training places”.

Whereas on 12 August the younger applicant was sent:

“Due to holiday commitments, we are unable to offer you an interview until early

September. If this is of interest to you, please contact … to arrange an appointment”.

It follows that, in most cases, a rejected applicant would not be aware that they had

incurred discriminatory treatment and, even if they did suspect it, they would lack

evidence to demonstrate, before a legal tribunal, a prima facie case of discrimination.

This is acknowledged, inter alia, by the OECD; “… age discrimination legislation

20

may not be very effective since it is often easier to prove discrimination in dismissal

than hiring” (OECD 2004, p. 99). It was also an important conclusion of Adams in his

investigation of the impact of state age discrimination legislation in the USA. Using

interstate data from as far back as the 1960s he concluded - “One thing is clear,

however. There is no evidence that suggests there are positive effects for protected

workers. The stock of older workers that are new hires did not change” (Adams 2004,

p. 237). It follows that this is a particularly serious problem for policy to address,

especially in view of the rates of discrimination detected in our experiment for people

as young as 39 and 47. Nevertheless, the British age discrimination legislation of 2006

requires the complainant to prove facts from which the Employment Tribunal could

conclude that the respondent has discriminated.

Consequently a new approach is required and we recommend that, to strengthen the

effectiveness of age discrimination legislation, in respect of recruitment, any equal

opportunity, or human rights, commission should be charged to play an active,

investigative role in the recruitment process; that is, it should have power to conduct

random audits of hiring and personnel practices. If employers were required to keep

all records of job applications for a period of twelve months, and obliged to justify

decisions on short-listing for interview and final choice of candidate, in the event of

random audit, it would create pressure for scrupulousness in the hiring decision. An

appropriate analogy can be drawn here with the capital market. Public corporations

have various duties with respect to reporting to shareholders, potential shareholders

and the business community at large. They are also subject to independent financial

audit, and they are usually required to satisfy an independent commission about

various aspects of their financial activities. In effect, capitalist economies provide a

21

range of regulations and checks to protect the owners of financial capital against

unscrupulous practices and guard against the waste of this resource. Therefore it

seems entirely appropriate that similar protection be afforded the owners of human

capital, and that steps be taken to prevent it being wasted through employers using

screening devices, such as race, sex or age, for purposes unrelated to job performance.

Barbara Bergmann has also advocated a similar policy (see Bergmann 1986, p. 158).

A complementary recommendation is that the approach to combating age

discrimination in recruitment should be one of affirmative action. Affirmative action

inevitably invokes fierce controversy and opposition from the privileged group –

usually white, middle-class, “prime-age”, protestant men.

In an investigative, or audit strategy, we recommend that employers should be

required to justify why appropriately-qualified post-40/post-50/post-60 applicants

have not been appointed. If the proportion of post/40/50/60 appointments is less than

(say) 66 percent of the proportion of appropriately-qualified post40/50/60 applications

then the employer should be required to review personnel policies and an auditor

would be involved in future selection procedures until significant improvement

occurred. In other words we are recommending a form of affirmative action for the

middle-aged and elderly. As Bergmann has explained - “The selection process often

does have important subjective elements, allowing plenty of leeway for making

mistakes as well as for decision-makers attitudes about race and gender to influence

outcomes. Thus, it is wrong to assume that the candidate chosen in the absence of

affirmative action is always or almost always better than all those sent away”

(Bergmann 1986, p. 104).

22

Age-based affirmative action would not incur some of the opposition traditionally

directed at race or sex-based affirmative action. We are only recommending that older

applicants be given proportionate treatment in jobs which they have already been

doing, or perhaps, at most, one step up in the hierarchy. The charge that they “only got

the job because of their age” could not hold up: they have already demonstrated they

have done the job. The other side of this coin is that the aged will not react, as some

blacks and females do, in opposing affirmative action, because of their concern that it

suggests they did not get the job on their merits. In this case their “merits” have

previously passed muster.

Another significant distinction in respect of age-based affirmative action is that,

whereas whites never become black, and only rarely do males become female, the

young do become old. In other words we should expect lesser hostility from the

“majority” group, as in this case they stand to benefit in their turn. Moreover in the

current demographic environment the alternative to ensuring a fair employment deal

for “older” workers are increased taxes to finance the growing pension bill.

23

Table 1: Results for the Age Discrimination Tests 1

Occupation

2

Location of test

3

Neither invited

4

Usable tests

5

Equal treatment

6

Discrimination against older

7

Discrimination against younger

8

Net Discriminationa

Graduate Total (number) Percent

England

373

47 100

15 31.9

30 63.8

2 4.3

28 59.6***

Retail Manager Total (number) Percent

London

273

27 100

3 11.1

8 29.6

16 59.3

-8 -29.6*

Waiter Total (number) Percent Total London (number) Percent Total Rest of England (number) Percent

England London

390

80 100 22 100 58 100

11 13.8

3 13.6 8 13.7

46 57.5 17 77.3 29 50

23 28.8 2 9.1 21 36.2

23 28.8*** 15 68.2*** 8 13.8

Note 1: Chi-squared tests were conducted on the response rates and the results are indicated in column 8 - * significant at the 0.05 level; ** significant at the 0.01 level; *** significant at the 0.001 level. a. A negative value indicates discrimination against the younger applicant.

24

Table 2: Results for the UK Sex and Race Discrimination Tests

Occupation Study Location of test Test on basis of Net Discrimination

percent Chartered accountant

Riach and Rich (2006)

England

Sex

-20.0*a

Computer analyst programmer

Riach and Rich (2006)

London and South East

Sex

-35.3**a

Engineer

Riach and Rich (2006)

London, South, South-East, Home Counties

Sex

23.1*

Secretary

Riach and Rich (2006)

London

Sex

-43.1***a

Accountant, Electronics Engineer, Sales representative, Secretary

Jowell and Prescott-Clarke (1970)

England

Race Asian West Indian

50.0*** 11.0

Clerk, Sales Representative, Secretary, Shop assistant

Hubbuck and Carter (1980)

Nottingham

Race Asian West Indian

42.0*** 43.0***

Accountant, Clerical, Management Trainee, Salesman, Secretary

McIntosh and Smith (1974)

Birmingham London

Race Asian/West Indian

30.0***

Note 1: Chi-squared tests were conducted on the response rates and the results are indicated for net discrimination: * significant at the 0.05 level; ** significant at the 0.01 level; *** significant at the 0.001 level. a. A negative value indicates discrimination against the male applicant.

25

Table 3: Results for the Age Discrimination Tests for Graduates, for Firms Noted as “Investors in People”

1

Occupation

2

Location of test

3

Usable tests

4

Equal treatment

5

Discrimination against older

6

Discrimination against younger

7

Net Discrimination

Graduate Total (number) Percent

England

13 100

7 53.9

6 46.2

0 0.0

6 46.2**

Note 1: Chi-squared tests were conducted on the response rates and the results are indicated in column 8: * significant at the 0.05 level; ** significant at the 0.01 level

26

Appendix Two Résumés Used in Job Applications for Waiter Résumé A

Personal Profile of …

Personal details Born 13-4-1977 email Qualifications GCSE - 6 Subjects two of which were English and Mathematics NVQ3 in Hospitality Supervision NVQ2 in Food and Drink Services ‘Computer Usage and Fine Wines of the World’ studied in my course at College Employment 2001 – present: Senior Waiter at Restaurante Venezia in Tunbridge Wells.

1997- 2000: Waiter at a restaurant in Brighton, which I moved to in order to gain a

more responsible post, and to have experience of serving foreign food and fine wine.

1994 -1997: My first job was at a hotel in Brighton, where I initially served breakfast and afternoon tea, but was soon promoted to serving lunch and dinner. Sporting and cultural interests Mountain-biking and films.

27

RÉSUMÉ B RÉSUMÉ OF ……. ADDRESS EMAIL AGE 47 EDUCATION

5 “O” levels (including English and Maths) Evening classes at College - including food service, wine appreciation, restaurant management and computer skills. NVQ2 in Food and Drink Services and NVQ3 in Restaurant Supervision.

EXPERIENCE

After leaving school I had a variety of jobs such as coffee shop server and barman. I became a waiter in 1988 and since then have had a variety of jobs in hotels and restaurants, serving English, French, Italian and Spanish food. I began in hotels in Torquay, Bath and Burford. My duties included, preparing the dining room and serving breakfast, lunch and dinner. I also helped with room service. Subsequently I moved to more senior and responsible jobs in restaurants in London, Oxford and Stratford-upon-Avon, serving food and wine. For the last three years I have been employed at “Claudettes” in Guilford as Deputy Head Waiter.

INTERESTS

I play competitive Squash and I enjoy restoring classic motor cycles. Also I enjoy using the internet.

28

References Adams, Scott J. (2004). ‘Age Discrimination Legislation and the Employment of

Older Workers’, Labour Economics, Vol. 11, No.2, pp. 219-241.

The B.E. Journal of Economic Analysis and Policy, (2006), Vol. 6, Issue 2.

www.bepress.com/bejeap/advances/6,issue2

Arrow, Kenneth (1973). ‘The Theory of Discrimination, in Ashenfelter, Orley and

Rees, Albert (eds.), Discrimination in Labor Markets, Princeton University

Press, Princeton, pp. 3-33.

Becker, Gary (1971). The Economics of Discrimination, 2nd edition. University of

Chicago Press, Chicago.

Bendick Jnr. Mark, Jackson, Charles and Romero, J. H. (1996). ‘Employment

Discrimination Against Older Workers: An Experimental Study of Hiring

Practices’, Journal of Aging and Social Policy, Vol. 8, No. 4, pp. 25-46.

Bendick Jnr. Mark, Brown, Lauren and Wall, Kennington (1999). ‘No Foot in the

Door: An Experimental Study of Employment Discrimination Against Older

Workers’, Journal of Aging and Social Policy, Vol. 10, No. 4, pp. 5-23.

Bergmann, Barbara (1986). The Economic Emergence of Women. Basic Books, New

York.

29

Blinder, Alan (1973). ‘Wage Discrimination: Reduced Form and Structural

Estimates’, Journal of Human Resources, Vol. 8, pp. 436-455.

Bovenkerk, Frank (1992). Testing Discrimination in Natural Experiments: A Manual

for International Comparative Research on Discrimination on the Grounds of

‘Race’ and Ethnic Origin. International Labour Office, Geneva.

Brown, Colin and Gay, Pat (1985). Racial Discrimination 17 Years After the Act.

Policy Studies Institute, London.

Census 2001. http://www.statistics.gov.uk/census2001/

Daniel, Kirsten and Heywood, John S. (2007). ‘The Determinants of Hiring Older

Workers: UK Evidence’, Labour Economics, Vol.14, No. 1, pp. 35-51.

Daniel, W. W. (1968). Racial Discrimination in England. Penguin Books, England.

Darity Jr., William A. and Mason, Patrick L. (1998). ‘Evidence on Discrimination:

Codes of Color, Codes of Gender’, Journal of Economic Perspectives, Vol. 12,

No. 2, pp. 63-90.

Department for Work and Pensions (2002). Simplicity, Security and Choice: Working

and Saving for Retirement. HMSO, London.

30

Fidell, L. S. (1970). ‘Empirical Verification of Sex Discrimination in Hiring Practices

in Psychology’, American Psychologist’, Vol. 25, December, pp. 1094-98.

Goldberg, Andreas, Mourinho, Dora and Kulke, Ursula (1996). Labour Market

Discrimination Against Foreign Workers in Germany. International Migration

Papers 7. International Labour Office, Geneva.

Gunderson, Morley (1989). ‘Male-Female Wage Differentials and Policy Responses’,

Journal of Economic Literature, Vol. 27, pp. 46-72.

Higher Education Statistics (2002/03). Statistics on Students in Higher Education in

the United Kingdom. Unpublished data. Higher Education Statistics Agency

(HESA), Cheltenham, UK.

Hubbuck, Jim and Carter, Simon (1980). Half a Chance? A Report on Job

Discrimination against Young Blacks in Nottingham. Commission for Racial

Equality, London.

Investors in People. 2005. http://www.iipuk.co.uk

Jowell, Roger and Prescott-Clarke, Patricia (1970). ‘Racial Discrimination and White-

Collar Workers in Britain’, Race, Vol. 11, No. 4, pp. 397-417.

31

La Piere, Richard (1934). ‘Attitudes vs Actions’, Social Forces, Vol. 13, No. 2, pp.

230-237.

Labour Market Statistics (2005). http://www.statistics.gov.uk/

List, John (2004). ‘The Nature and Extent of Discrimination in the Marketplace:

Evidence from the Field’, Quarterly Journal of Economics, Vol.119, No. 1, pp.

49-89.

McIntosh, Neil and Smith, David (1974). ‘The Extent of Racial Discrimination’,

Political and Economic Planning Broadsheet no.547, Political and Economic

Planning, London.

Oaxaca, Ronald (1973). ‘Male-Female Wage Differentials in Urban Labor Markets’,

International Economic Review, Vol. 14, No. 3, pp. 693-703.

OECD (2004). Ageing and Employment Policy, United Kingdom. Paris, OECD.

OECD (1992). The Job Study, Part 11 – The Adjustment Potential of the Labour

Market. OECD, Paris.

Orwell, George (1940). Down and Out in Paris and London. Penguin,

Harmondsworth, UK.

32

Parliamentary Debates (1968). Volume 763, column number 78, 23 April. HMSO,

London.

Parliamentary Debates (1967). Volume 754, column number 452, 15 November.

HMSO, London.

Penna Sanders and Sidney (2002). ‘Generation Flex to the Rescue?’

http://www.management-issues.com.

Phelps, Edmund (1972). ‘The Statistical Theory of Racism and Sexism’, American

Economic Review, Vol. 62, No. 4, pp. 659-661.

Purcell, Kate, Wilton, Nick and Elias, Peter (2003). Older and Wiser? Age and

Experience in the Graduate Labour Market. Researching Graduate Careers

Seven Years On. Research paper no. 2. Warwick University, Warwick Institute

for Employment Research.

Riach, Peter A. and Rich, Judith (2004a). ‘Deceptive Field Experiments of

Discrimination: Are They Ethical?’, KYKLOS, Vol. 57, Issue 3, pp. 457-470.

Riach, Peter A. and Rich, Judith (2004b). ‘Fishing for Discrimination’, Review of

Social Economy, Vol. 61, No. 4, pp. 465-486.

Riach, Peter A. and Rich, Judith (2002). ‘Field Experiments of Discrimination in the

Market Place’, Economic Journal, Vol. 112, No. 483, pp. F480-F518.

33

Riach, Peter A. and Rich, Judith (1987). ‘Testing for Sexual Discrimination in the

Labour Market’, Australian Economic Papers, Vol. 26, No. , pp. 165 – 178.

Smith, Zadie (2000). White Teeth. Hamish Hamilton, London. Third Age Employment Network (2003). Experience of Age Discrimination: The

Evidence. February, http://www.taen.org.uk

Warr, P. (1994). ‘Age and Job Performance’, In: Snel, Jan and Cremer, Roel (eds.),

Work and Aging: A European Perspective, Taylor and Francis, London, pp.

309-322.

34